Through the work of The Cancer Genome Atlas (TCGA), researchers have confirmed the most common form of malignant brain cancer in adults, glioblastoma multiforme (GBM), isn’t one disease as once believed but is, instead, four distinct molecular subtypes.

 

Identifying these different subtypes and understanding the specific gene expressions in each could lead to new treatment strategies. Using integrative genetic analysis, the team found defining characteristics that linked each type of GBM tumor to a specific … and different … molecular process. The team’s findings were published in the Jan. 19, 2010 edition of Cancer Cell.

 

Lead author D. Neil Hayes, MD, associate professor at the University of North Carolina — Chapel Hill, noted, “This was collaborative from start to finish. The credit goes to all the investigators and institutions involved with TCGA and also the patients and institutions that donated the tissue.”

 

Although suspected from previous research, the TCGA effort produced solid evidence that GBM is indeed made up of specific subtypes: Proneural, Neural, Classical and Mesenchymal. Using a high quality set with about 250 brain cancer samples, Hayes said the team used parallel technologies … primarily gene sequencing and copy number analysis … to better understand the subtypes and create a “parts” list in 2008. This list, which included sequence data on more than 600 genes, showed TP53 and EGFR mutated very frequently in GBM. However, Hayes noted, it didn’t show how the various parts worked together to form the various types of GBM.

 

“We took that parts list and now we’re putting together a blueprint,” he said of the research published in January. “We showed the cancer … at least glioblastoma … is not a random grab bag of events.” He continued, “We discovered a bundle of events that unequivocally occur almost exclusively within a subtype. These are critical events in the history of the tumor’s development and spread, and evidence is increasing that they may relate to the initial formation of the tumors.”

 

Hayes noted that although TP53 frequently mutates in GBM, this isn’t the case in Classical GBM. For this subtype, the cancer occurs because of a mutation and alteration in EGFR along with the loss of another gene, P16. Conversely, in Proneural, TP53 is altered in almost every sample along with PGDFRA and IDH1.

 

Knowing which cells to target for each of the subtypes could potentially change treatment protocols. The study showed the subtypes responded differently to a regimen of aggressive chemotherapy and radiation that is currently being used. Patients with the Classical subtype succumbed to the disease at a rate approximately 50 percent slower when treated aggressively. Neural and Mesenchymal also enjoyed some survival benefit … although to a significantly lesser degree … with aggressive chemotherapy and radiation. Proneural, which is the subtype that tends to do the best overall, saw no added benefit from the more rigorous course of treatment.

 

While these findings don’t affect current clinical practice, TCGA investigators believe their research provides a framework for further discovery — ultimately leading to personalized approaches for GBM based on a patient’s specific genomic alteration. Hayes said the implications of this research leads to three key ‘next steps.’

 

“Number one, it raises a question for basic scientists … why? It’s clearly not by random chance so we need to find out why this happens,” he said of the cascade of mutations and alterations that lead to each form of GBM. “Number two, it allows us to tie a specific subtype of cancer to a specific model of cancer.” Hayes continued, “We’re looking to distinguish types of brain cancer in the same way we distinguish types of leukemia.” The third step, he said, is to design clinical trials aligned to the specific subtype to improve patient diagnoses and outcomes.